Thermodynamic Model of the Glass Transition Behavior for Miscible

Jan 6, 2006 - Dong Hoon Lee , Yong Soo Kang , Jong Hak Kim , Jong Hak Kim ... Yong Woo Kim , Jung Tae Park , Joo Hwan Koh , Byoung Ryul Min , Jong ...
0 downloads 0 Views 62KB Size
Macromolecules 2006, 39, 1297-1299

1297

Notes w1 w2 1 ) + Tg Tg1 Tg2

Thermodynamic Model of the Glass Transition Behavior for Miscible Polymer Blends Jong Hak Kim,*,† Byoung Ryul Min,† and Yong Soo Kang‡ Department of Chemical Engineering, Yonsei UniVersity, 134 Shinchon-dong, Seodaemun-gu, Seoul 120-749, South Korea, and DiVision of Chemical Engineering, Hanyang UniVersity, 17 Haengdang-dong, Seongdong-gu, Seoul 133-791, South Korea ReceiVed NoVember 15, 2005 ReVised Manuscript ReceiVed December 12, 2005 Introduction Polymer blends have received much attention since blending is a simple, effective approach to develop new materials exhibiting combinations of properties that cannot be obtained by individual polymers.1,2 Miscibility between two components is governed by the thermodynamics represented by the Gibbs free energy of mixing (∆Gmix ) ∆Hmix - T∆Smix). The mixtures are miscible when the value of ∆Gmix is negative, i.e., low value of ∆Hmix and high value of ∆Smix. Immiscibility is a rule in polymer blends because both the ∆Hmix and the ∆Smix are unfavorable. The ∆Smix is unfavorable because there are few molecules of large molecular weight per unit volume. The ∆Hmix is also unfavorable because the van der Waals dispersion force is of the form of -AB/r6 (A and B are the Hamaker constants)3 so that compared to the reference state we have for ∆Hmix

∆Hmix ) -2

AB AA BB (A - B) + 6 + 6 ) r6 r r r6

2

(1)

which is always positive. However, when the specific interactions such as hydrogen bonding, dipole-dipole interaction, or ionic interaction are established, the miscibility between two polymer mixtures can be achieved. The glass transition temperature (Tg) reflects the molecular rearrangement rate in supercooled liquid and thus is one of the most important factors among the many transitions and relaxations in polymers. Measurement of Tg is one of the easiest determinations of whether they are miscible or immiscible in polymer mixtures.4-6 A single Tg between individual two polymers appears for miscible blends whereas two Tgs are shown for immiscible blends. When Tg of miscible polymer blends shows the linear relationship against the polymer composition, the value of Tg is adequately expressed by the widely used Fox equation.7-13 †

Yonsei University. Hanyang University. * To whom correspondence should be addressed: Tel +82-2-2123-5757; Fax +82-2-312-6401; e-mail [email protected]. ‡

(2)

where w1 and w2 are the weight fractions of components 1 and 2 with Tg1 and Tg2, respectively. This equation is also used to predict the Tg of copolymers. For some miscible polymer blends interacting with strong specific intermolecular and intramolecular forces, it has been reported that the compositional dependence of Tg shows a maximum behavior, exhibiting a positive deviation from linearity with blend composition.14,15 This unusual behavior was ascribed to the significant hydrogen-bonding interactions between two polymers in those papers. The Tg of such polymer blends is of special interest from a practical and academic point of view.16 In this study, we propose a simple, molecular thermodynamic model of Tg in polymer blends based on the configurational entropy model17-20 and Flory-Huggins theory21 to predict the glass transition behavior in polymer blends with and without specific interaction between polymers. Thus, the present Tg model encompasses not only the maximum behavior in the polymer blends with specific interactions but also the simple linear relationship against composition. Model Development Gibbs and DiMarzio suggested that glass formation is a result of the system’s loss of configurational entropy (Sc).17

Sc ) Sliquid - Sglass

(3)

where Sliquid and Sglass indicate the configurational entropies of the liquid and glass states, respectively. It has been assumed that Sglass(n1,0,T) ) Sglass(n1,n2,T) ) 0 and that ∆Cp is independent of temperature and composition.17,19 The change in the glass transition temperature produced by blending polymer 1 and the polymer 2 can be described by19,22,23

ln

( )

Tg12 1 )[S (n ,n ,T) - Sc(n1,0,T)] Tg1 ∆Cp c 1 2

(4)

where Tg1 and Tg12 are the Tgs of the pure polymer 1 and of the blend of polymer 1 and polymer 2, respectively. ∆Cp is the difference in the heat capacity between the supercooled liquid and the glass. n1 and n2 are the numbers of molecules of polymer 1 and polymer 2, respectively. Sc for polymer blends consists of the disorientation entropies of polymer 1 (Sdis-1) and polymer 2 (Sdis-2), the mixing entropy (Smix-12), and the specific interaction entropy (Sspe-12).

Sc(n1,n2,T) ) Sdis-1 + Sdis-2 + Smix-12 + Sspe-12

(5)

Sc(n1,0,T) ) Sdis-1

(6)

Substitution of eqs 5 and 6 into eq 4 yields

10.1021/ma052436a CCC: $33.50 © 2006 American Chemical Society Published on Web 01/06/2006

1298

Notes

ln

Macromolecules, Vol. 39, No. 3, 2006

( )

Tg12 1 )[S + Smix-12 + Sspe-12] Tg1 ∆Cp dis-2

(7)

Since the specific interactions such as hydrogen bonding, dipole-dipole interaction, or ionic interaction between two polymers reduce the overall entropy, the value of Sspe-12 is negative. However, other entropy terms increase the overall entropy, thereby being positive values. The entropy terms are expressed as follows:22-25

[

Sdis-2 ) kBn2 ln r2 + (r2 - 1) ln

(z -e 1)]

(8)

Smix-12 ) -kB[n1ln φ1 + n2 ln φ2] Sspe-12 ) γspekB ln

(9)

( )(

φ2 z - 1 φ1 ln φ1 + ln φ2 e r1 r2

)

(10)

where kB is the Boltzmann constant. φ1 ) r1n1/(r1n1 + r2n2) and φ2 ) r2n2/(r1n1 + r2n2) are the volume fractions of components 1 and 2, respectively. r1 ) V1/V0 and r2 ) V2/V0, where V1 and V2 are the molar volumes of components 1 and 2, respectively, and V0 is the unit lattice volume. z is the lattice coordination number. γspe is a proportionality constant representing the specific interaction such as hydrogen bonding, dipole-dipole interaction, or ionic interaction between two polymers. Combination of eqs 7-10 yields the following equation:

ln

( ) [(

))(

)

φ2 Tg12 z - 1 φ1 ) β 1 - γspe ln ln φ1 + ln φ2 Tg1 e r1 r2 φ2 z-1 ln r2 + (r2 - 1) ln (11) r2 e

(

[

)]]

(

where β ) zR/(M1u∆Cpp). R, M1u, and ∆Cpp are the gas constant, the molecular weight of the repeat unit, and the isobaric specific heat of polymer 1, respectively. When we limit the value of φ1 to be zero, eq 11 would be expressed as follows:

( )

ln

[

)]

Tg2 z-1 β ) - ln r2 + (r2 - 1) ln Tg1 r2 e

(

(12)

Equations 11 and 12 can be combined to provide a configurational entropy theory for the glass transitions of binary polymer blends.

ln

( ) (

))( r

Tg12 z-1 ) β 1 - γspe ln Tg1 e

(

φ1 1

ln φ1 +

)

φ2 ln φ2 + r2 Tg2 (13) φ2 ln Tg1

( )

From this model, we can predict Tg for the polymer blends with various compositions. Results and Discussion Figure 1 shows the theoretical prediction of Tg calculated from eq 13 for polymer blends (polymer 1/polymer 2) with various specific interaction parameters (γspe) as a function of volume fraction of polymer 2. To simplify the system, the physical parameters for theoretical calculation are assumed to have the following values: β ) 1.0, z ) 4.0, r1 ) 100, r2 ) 100. When there is no specific interaction between two polymers (γspe )

Figure 1. Theoretical prediction of Tg as a function of volume fraction of polymer 1 with varying specific interaction parameter (γspe) for blends of polymer 1 and polymer 2. The physical parameters used are at β ) 1.0, z ) 4.0, r1 ) 100, and r2 ) 100.

0), the Tg shows a linear relationship against the composition of polymer blends, which is consistent with the Fox equation. As γspe values increase, the maximum behavior of Tg becomes more prominent with larger positive deviation, representing the stronger specific interaction between two polymers. When the difference of Tg between two polymers is small, the positive deviation becomes more noticeable. Figure 2 presents the comparison of the theoretical predictions with experimental glass transition temperature for blends of poly(vinylpyrrolidone) (PVP) and poly(vinylphenol) (PVPh). Kuo and Chang14 investigated the hydrogen-bonding interaction and miscibility behavior of blends of PVP and PVPh by differential scanning calorimetry (DSC). They observed that this blend is able to form a miscible phase due to the formation of inter-hydrogen bonding between the carbonyl of PVP and the hydroxyl of PVPh. Interestingly enough, a single Tg higher than that of either individual polymer was observed, presumably due to the strong interaction between two polymers. As seen in the figure, the new model eq 13 predicts accurately the Tg values from the experimental results, but the Fox equation does not. The value of γspe of eq 13 was obtained by nonlinear regression. The higher value of γspe () 288.1) for blends of PVP and PVPh represents the strong specific hydrogen-bonding interaction between two polymers, which is coincident with the previous experimental results of FT-IR and solid-state NMR.14 The new thermodynamic model of Tg was also applied to the system exhibiting a linear relationship against compositions of polymer blends. Balsamo et al.13 prepared blends of polycarbonate (PC) and poly(-caprolactone) (PCL) in a wide

Macromolecules, Vol. 39, No. 3, 2006

Figure 2. Comparison of the theoretical predictions with experimental glass transition temperature for blends of PVP(1)/PVPh(2). Open squares are experimental data by Kuo and Chang.14 The solid line is calculated from eq 13, where γspe ) 288.1, r1 ) 88.8, r2 ) 103.4, z ) 4, M1u ) 111 g/mol, and ∆Cpp ) 740.5 J/(kg K).

Notes

1299

equation based on conformational entropy changes. In that study, the specific interactions were not considered, and thus they failed to predict the experimental Tg values well. This presents the importance of a physical parameter γspe in the current thermodynamic Tg model. In conclusion, a new mathematical model was derived on the basis of configurational entropy and Flory-Huggins theory to predict the Tg of miscible polymer blends. The new model adequately predicts the Tg dependence on blend composition for the cases of (1) maximum behavior with positive deviation and (2) simple linear additivity obeying the Fox equation. More importantly, the current Tg model does not include other adjustable parameters except γspe and can be applied to any miscible binary polymer blends with hydrogen bonding, dipoledipole interaction, or ionic interaction. From this model, we can directly compare the strength of specific interaction in each system by the value of γspe; i.e., the higher value of γspe for polymer blends represents the stronger specific interaction and vice versa. Acknowledgment. This work was supported by Yonsei University Research Fund of 2005. References and Notes (1) (2) (3) (4) (5) (6) (7) (8) (9)

Figure 3. Comparison of the theoretical predictions with experimental glass transition temperature for blends of PC(1)/PCL(2). Open squares are experimental data by Balsamo et al.13 New model equation is calculated from eq 13, where γspe ) 2.74, r1 ) 222.8, r2 ) 100.0, z ) 4.0, M1u ) 254 g/mol, and ∆Cpp ) 388.2 J/(kg K).

composition range and investigated the thermal behavior of the blends via DSC. The blends were miscible in the entire composition range, showing the single Tg values. The compositional variation of Tg exhibits a simple, linear relationship, which is adequately described by the Fox eq 2,7,8 as seen in Figure 3. The new model eq 13 also predicts the experimental Tg values very accurately, comparable to the Fox equation. The lower value of γspe ()2.74) for blends of PC/PCL relative to the blends of PVP/PVPh demonstrates the weak specific interaction between two polymers in the former. Schneider and Di Marzio26 reported on the comparison of experimental Tg values for miscible polymer blends with an

(10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26)

Ruzette, A.-V. G.; Mayes, A. M. Macromolecules 2001, 34, 1894. Gonzalez-Leon, J. A.; Mayes, A. M. Macromolecules 2003, 36, 2508. Hamaker, H. C. Physica 1937, 4, 1058. Kuo, S.-W.; Lin, C.-L.; Chang, F.-C. Polymer 2002, 43, 3943. Chiu, C.-Y.; Chen, H.-W.; Kuo, S.-W.; Huang, C.-F.; Chang, F.-C. Macromolecules 2004, 37, 8424. Takeda, K.; Murata, K.; Yamashita, S. J. Non-Cryst. Solids 1998, 231, 273. Fox, T. G.; Loshaek, S. J. Polym. Sci. 1955, 15, 371. Fox, T. G.; Flory, P. J. J. Polym. Sci. 1954, 14, 315. Cowie, J. M. G.; Harris, S.; Ribells, J. L. G.; Meseguer, J. M.; Romero, F.; Torregrosa, C. Macromolecules 1999, 32, 4430. Shi, Y.; Jabarin, S. A. J. Appl. Polym. Sci. 2001, 81, 11. Ishida, H.; Lee, Y. H. J. Appl. Polym. Sci. 2001, 81, 1021. Park, M. S.; Kim, J. K. J. Polym. Sci., Polym. Phys. 2002, 40, 1673. Balsamo, V.; Calzadilla, N.; Mora, G.; Mu¨ller, A. J. J. Polym. Sci., Polym. Phys. 2001, 39, 771. Kuo, S.-W.; Chang, F.-C. Macromolecules 2001, 34, 5224. Takeda, K.; Murata, K.; Yamashita, S. J. Phys. Chem. B 1999, 103, 3457. Lu, X.; Weiss, R. A. Macromolecules 1992, 25, 3242. Gibbs, J. H.; DiMarzio, E. A. J. Chem. Phys. 1958, 28, 373. DiMarzio, E. A.; Gibbs, J. H. J. Polym. Sci. 1963, A1, 1417. Chow, T. S. Macromolecules 1980, 13, 362. Gordon, J. M.; Rouse, G. B.; Gibbs, J. H., Jr.; Risen, W. M. J. Chem. Phys. 1977, 66, 4971. Flory, P. J. Principles of Polymer Chemistry; Cornell University Press: Ithaca, NY, 1953. Kim, J. Y.; Hong, S. U.; Won, J.; Kang, Y. S. Macromolecules 2000, 33, 3161. Kim, J. H.; Min, B. R.; Won, J.; Kang, Y. S. J. Phys. Chem. B 2003, 107, 5901. Kim, J. Y.; Bae, Y. C. J. Appl. Polym. Sci. 1999, 73, 1891. Kim, J. Y.; Bae, Y. C. Fluid Phase Equilib. 1999, 163, 291. Schneider, H. A.; Di Marzio, E. A. Polymer 1992, 33, 3453.

MA052436A